SPRINGLab/IndicTTS_Marathi

Marathi Indic TTS Dataset This dataset is derived from the Indic TTS Database project, specifically using the Marathi monolingual recordings from both male and female speakers. The dataset contains high-quality speech recordings with corresponding text transcriptions, making it suitable for text-to-speech (TTS) research and development. Dataset Details Language: Marathi Total Duration: ~10.33 hours (Male: 5.16 hours, Female: 5.18 hours) Audio Format: WAV Sampling… See the full description on the dataset page: https://huggingface.co/datasets/SPRINGLab/IndicTTS_Marathi.